Underwater 3D image reconstruction utilizing triple wavelength dispersion and camera system thereof

1. A camera system comprising:

a projector selectively operable to project a pattern of electromagnetic radiation toward a target object, the electromagnetic radiation comprising at least three different wavelengths including a first wavelength, a second wavelength, and a third wavelength;

a sensor directed toward the target object to receive reflected electromagnetic radiation from the target object;

a storage device; and

one or more processors coupled to the projector, sensor, and storage device;

wherein, by the one or more processors executing software loaded from the storage device, the one or more processors are operable to:

send one or more control signals to the projector causing the projector to project the pattern of electromagnetic radiation toward the target object at each of the first, second and third wavelengths;

store into the storage device a plurality of image data received from the sensor, the image data corresponding to the reflected electromagnetic radiation detected by the sensor at each of the first, second, and third wavelengths;

process the image data to compute a refractive normal according to a wavelength dispersion represented by differences in the image data at each of the first, second, and third wavelengths;

process the image data to compute an interface distance corresponding to a distance from a center point of the sensor to a first refractive interface nearest the sensor according to the refractive normal;

generate a 3D representation of the target object by back projecting each pixel of the image data at the first, second, and third wavelengths in order to determine an object point location according to the refractive normal and interface distance; and

store the 3D representation of the target object in the storage device.

2. The camera system of claim 1 , wherein the electromagnetic radiation has wavelengths within 400-700 nanometers.

3. The camera system of claim 1 , wherein each of the first wavelength, second wavelength, and third wavelength are separated by at least 100 nanometers.

4. The camera system of claim 1 , wherein the projector comprises at least three light emitting diodes each respectively emitting one of the first, second, and third wavelengths of electromagnetic radiation.

5. The camera system of claim 1 , wherein the first wavelength corresponds to red light, the second wavelength corresponds to green light, and the third wavelength corresponds to blue light.

6. The camera system of claim 1 , wherein the one or more processors are operable to:

send the control signals to the projector thereby causing the projector to project three separate instances of the pattern of electromagnetic radiation toward the target object at different times, a first instance of the pattern being at the first wavelength, a second instance being at the second wavelength, and a third instance being at the third wavelength; and

include three corresponding images in the image data, a first image corresponding to reflected electromagnetic radiation at the first wavelength, a second image corresponding to reflected electromagnetic radiation at the second wavelength, and a third image corresponding to reflected electromagnetic radiation at the third wavelength.

7. The camera system of claim 1 , wherein the projector is operable to project the pattern of electromagnetic radiation toward the target object at each of the first, second and third wavelengths at a same time.

8. The camera system of claim 1 , wherein the one or more processors is further operable to apply a refinement process to the 3D representation of the target object by:

reprojecting all points of the 3D representation of the target object back to the image data, determining a reprojection error, and

utilizing the reprojection error as a cost function to refine each point of the 3D representation of the target object.

9. The camera system of claim 1 , wherein the sensor is mono.

10. The camera system of claim 1 , wherein the sensor is color.

11. An autonomous underwater vehicle (AUV) incorporating the camera system of claim 1 .

12. A remotely operated vehicle (ROV) incorporating the camera system of claim 1 .

13. A method comprising:

projecting a pattern of electromagnetic radiation toward a target object, the electromagnetic radiation comprising at least three different wavelengths including a first wavelength, a second wavelength, and a third wavelength;

receiving reflected electromagnetic radiation from the target object at a sensor;

storing a plurality of image data, the image data corresponding to the reflected electromagnetic radiation detected at each of the first, second, and third wavelengths;

processing the image data to compute a refractive normal according to a wavelength dispersion represented by differences in the image data at each of the first, second, and third wavelengths;

processing the image data to compute an interface distance corresponding to a distance from a center point of the sensor to a first refractive interface nearest the sensor according to the refractive normal;

generating a 3D representation of the target object by back projecting each pixel of the image data at the first, second, and third wavelengths in order to determine an object point location according to the refractive normal and interface distance; and

storing the 3D representation of the target object in a storage device.

14. The method of claim 13 , wherein the electromagnetic radiation has wavelengths within 400-700 nanometers.

15. The method of claim 13 , wherein each of the first wavelength, second wavelength, and third wavelength are separated by at least 100 nanometers.

16. The method of claim 13 , further comprising projecting the pattern of electromagnetic radiation toward the target object utilizing at least three light emitting diodes each respectively emitting one of the first, second, and third wavelengths of electromagnetic radiation.

17. The method of claim 13 , wherein the first wavelength corresponds to red light, the second wavelength corresponds to green light, and the third wavelength corresponds to blue light.

18. The method of claim 13 , further comprising

projecting three separate instances of the pattern of electromagnetic radiation toward the target object at different times, a first instance of the pattern being at the first wavelength, a second instance being at the second wavelength, and a third instance being at the third wavelength; and

including three corresponding images in the image data, a first image corresponding to reflected electromagnetic radiation at the first wavelength, a second image corresponding to reflected electromagnetic radiation at the second wavelength, and a third image corresponding to reflected electromagnetic radiation at the third wavelength.

19. The method of claim 13 , further comprising projecting the pattern of electromagnetic radiation toward the target object at each of the first, second and third wavelengths at a same time.

20. The method of claim 13 , further comprising applying a refinement process to the 3D representation of the target object by:

reprojecting all points of the 3D representation of the target object back to the image data,

determining a reprojection error, and

utilizing the reprojection error as a cost function to refine each point of the 3D representation of the target object.